Profile building with respect to chosen subfield of application
-
How do you build your profile for Masters and PhD programs, especially during your sophomore and junior years while you are preparing to apply to graduate school. Secondly, how should we decide which specific field to apply in? At least many students in LUMS have the perception that Machine Learning (ML) is particularly hard to get into, so many people change their fields and actually turn to Computer Networks or to HCI because of a relatively higher chance of getting accepted. Further, many believe it is the hardest to publish a paper in ML, and since it is may be challenging to get into a PhD program without a research paper, they end up changing their desired subfield of study.
-
Profile building varies significantly for both Master's and PhD applicants. For PhD, it is highly valuable to have independent research experience. It signals to the admissions committee that you understand the ups and down of the research process and the uncertainty inherent in research. Without this experience, the admissions committee may find it difficult to assess how well you can manage the transition to the demanding requirements of a PhD program and your potential to succeed.
Broadly speaking, you should try to strive to do well in your courses, especially in subjects relevant to your intended field of graduate study. If you are applying to Computer Science (CS) programs then having good grades in CS, Math and Engineering courses matter more than others because it demonstrates not only your background knowledge but also your ability to grasp complex concepts, and to a certain degree your commitment, work ethic and discipline. Additionally, having independent work experience can significantly enhance your profile. Whether through research projects, internships or even your course projects. This will demonstrate to the admissions committee that you are prepared for graduate-level work and are capable of handling complex tasks independently, which is the norm in graduate school.
-
I would like to address the concern regarding applying to program specialities based on admit chances as opposed to applying to programs of interest. I personally think students should apply strictly to programs they are interested in and are actually passionate about. In our conversations with colleagues who also serve as admission committee members for some of the top competitive programs like CMU, we found that having publications is not a hard requirement, even for PhD programs. In fact they shared that many students apply with publications in subpar journals which do not necessarily have expert reviewers. Such publications are not impressive and end up having little value for the admissions committee. Unless the publications are in top and respectable venues, they are sometimes not even counted in the evaluation process. The admissions committee needs evidence that you have trained in handling complex problems and can dive deeper into the subject matter to pitch and implement a solution to a hard practical problem. This evidence can come from your letters of recommendation from supervisors who can speak to your relevant problem-solving abilities. Your letters of recommendations are looked at carefully to assess how you are perceived by an academic, and how they attest to your ability to succeed in the graduate program.
Regarding the ‘profile-building’ aspect, I personally think this notion can confuse you because it implies you need to perfect all elements of your application. This can lead to loss in the quality or substance of your unique selling points. I think your focus should be, starting in your undergraduate, to try to best learn in your foundational courses and build a strong basis. Then you can proceed with more advanced coursework in your areas of interest moving forward. For example if your probability and statistics courses particularly intrigued you, you can take Machine learning, deep learning and generative AI courses to further build and explore your interest. Showcase that you followed and pursued your interests and committed to them, which can be done through good grades and your letters of recommendation. In fact, many times students have to go through long interviews for PhDs and Masters with funding and may be asked to showcase their technical acumen based on their course learnings through programming exercises and technical questions. It is therefore vital to cover your bases and have strong foundational knowledge in your areas of interest.
-
Following Dr. Zafar's insights, I'd like to emphasize that all students in Computer Science or related fields should possess sound data skills. Virtually no fields today can operate without some reliance on data skills, whether directly or indirectly. Whether your focus is on networks, security, or human-computer interaction, data skills are now essential tools in both academic and industrial settings. Building a strong foundation in mathematics and analytical skills through courses in statistics, probability, and linear algebra will also significantly enhance your capabilities within your chosen specialties.
-
Regarding building your portfolio, I’d like to request you to first pick a sport. As an exercise or training that you do everyday. Now let me tell you why. If you pick a profession or field which requires you to sit everyday, enclosed in a space, sometimes up to 17-18 hours a day, your mental and physical capacities will begin to decline. Many people start developing mental health issues, they go into depression or develop physical issues. You might graduate with a 3.97 but would it have been worth it if you also graduate with a lifelong backbone problem? This is honestly the first piece of advice I give to all my students, and something me and my colleagues also practice. Pick a game or physical activity that you like and stick to it.
Secondly, I would like to emphasize that one thing that is very hard to recover or improve is your GPA. I am not saying that a high GPA is a hard prerequisite for admission. There are countless examples of students with very low GPAs that have made into some of the most brilliant institutes worldwide. Dr. Ihsan, Dr. Zafar and myself have advised and mentored several students who fall in this category. However, It does become very difficult with a low GPA. Students often have to do some outstanding work in other components to counteract a low GPA. While of course, GPAs are relative and in some universities it is much harder to get the same GPA, you should aim for a number close to 3.5 or above. This is a decent enough number to make you competitive and you should try to remain above this number. Even if you have a lower GPA right now, you can still touch this number or come close with requisite effort in your remaining courses.
Regarding the competition in ML: As someone who teaches and conducts workshops in the field of ML and generative AI, I want you to know that the industry is still in the ‘euphoria’ phase of it all. Many companies that make LLMs for example, are realizing that a primary bottleneck is going to be network bandwidth and there is going to be a subsequent boom in Network-centered solutions too. We are still in the phase where we do not fully realize the complete repercussions of these developments. Similarly, security is going to be a big issue moving forward. ChatGPT for example, sometimes shares URLs mistakenly which leave the user prone to comm-injection attacks.Furthermore, bots are only as good as the interface they are embedded into and its level of engagement with the user, meaning HCI is also going to be essential to this boom. So rest assured, all fields are going to remain relevant in their own unique way and you should follow your passion. Conduct a small mental exercise with me. If you were at a bus stop, and you needed to go to a particular destination but all buses leading to that destination were full, you wouldn’t pick just any other random bus now would you? In a very similar manner, please do not apply to programs you think are easier to get into, but are actually not passionately invested in. You are in your prime and at the peak of your learning ability. It would be a grave mistake to believe you can later on switch easily into your relevant field after you have graduated or settled into a specific industry. I believe that if getting admitted is harder now, it is going to be even more difficult to switch fields later on, given how competitive the arena is now. Follow what you like, and the path to success will follow you too. Be good at what you do, and you will always have plenty of opportunities in terms of admissions, funding and jobs.
